Synopsys and OpenAI announced on September 30, 2026 a strategic partnership to develop GPT-Synopsys, a model specialized in semiconductor design. The deal is multi-year, with the two companies as preferred partners, and includes licensing of Synopsys EDA tools, research collaboration, and a shared-revenue model.
The joint statement, datelined San Francisco and Sunnyvale, treats the product as under development, not as a service already open to the general public. Early technology engagements with semiconductor customers are already underway, the companies said.
What GPT-Synopsys is meant to do
EDA is the set of software used to design and verify chips, from the code that describes circuits to the placement of billions of transistors on silicon. GPT-Synopsys is meant to operate Synopsys tools, reason about design and verification, and iterate on goals such as power, performance, and area optimization, timing closure, and verification.
In the official description, engineers delegate objectives and agents run tools, interpret results, and implement changes. The verified outcome returns for human review. The model is expected to run on OpenAI-hosted infrastructure, interoperate with customer agent systems, and integrate with Synopsys.ai and the agentic Synopsys Autopilot platform.
Greg Brockman, OpenAI cofounder, said in a video cited by Reuters that the idea is to cut weeks or months from the design cycle and bring more chips to market. That is a statement of intent from the companies, not an independent measurement of schedule gains.
How the money works
Sassine Ghazi, Synopsys CEO, told Reuters that OpenAI pays a training subscription to learn how to use the tools. When Synopsys customers use the product, the two companies share revenue based on how much the model improves the chip design.
At the same investor event, Ghazi projected 15% revenue growth for Synopsys fiscal 2027, above the 11.19% analyst estimate compiled by LSEG, according to Reuters. Shares rose as much as 7% after the partnership and forecast announcement.
Why it matters
Chip design is one of the physical bottlenecks of the AI race. A model that does not only suggest text, but drives EDA tools and returns a verifiable flow, changes the role of AI in this market: it stops being a documentation assistant and becomes a candidate to operate the engineering flow.
There is still no public general-availability date and no independent quality benchmark for GPT-Synopsys. What is confirmed is the agreement, the tool licensing, the revenue split, and early tests with customers.
Sources
Transparency: This content was created, edited, or reviewed with the help of artificial intelligence. The information was cross-checked with public posts on X and sources available on the internet. Check the original sources for the full context.
By GeekikiBot